Unexpected bonding mode of the diboran(4)yl ligand: Combining the boryl motif with a dative Pt-B interaction

Article abstract of DOI:10.1002/anie.201102593

Catching the anchor: Although diboranes(4) commonly react by way of B-B bond activation, reaction of B₂Mes₂Br₂ with [Pt(PEt₃)₃] led to the selective oxidative addition of one B-Br bond; the resulting diboran(4)yl complex shows an unexpected dative Pt-B bonding interaction to the second boron center of the diboran(4)yl ligand. Oxidative addition of one B-Br bond in B₂(NMe₂) ₂Br₂ to [Pt(PiPr₃)₂] enabled the isolation of a diboran(4)yl species without any dative Pt-B interaction. Copyright

Full text of DOI:10.1002/anie.201102593

DOI: 10.1002/anie.201102593
Boron Ligands
Unexpected Bonding Mode of the Diboran(4)yl Ligand: Combining
ꢀ
the Boryl Motif with a Dative Pt B Interaction**
Holger Braunschweig,* Alexander Damme, and Thomas Kupfer
ꢀ
During the last 15 years, diboranes(4) have become increas-
ingly important as high-value reagents in the catalytic
functionalization of unsaturated organic substrates.[1] In this
regard, the commonly observed reactivity of diboranes(4)
towards the catalytic active species, which is usually a low-
whether a selective oxidative addition of B X bonds is
ꢀ
possible in the presence of a B B bond. Bearing the
experiences of Norman and co-workers in mind, we reasoned
that diboranes(4) containing more reactive B Br bonds might
be better suited for this purpose. This assumption proved to
be true and we were able to isolate trans-[(Et3P)2Pt(Br)-
{B(Mes)B(Mes)(Br)}] (1) and trans-[(iPr3P)2Pt(Br){B-
(NMe2)B(NMe2)(Br)}] (2), each prepared by oxidative addi-
tion of one B Br bond to the respective platinum precursors.
As discussed later on, the solid-state structure of 1 features a
hitherto unknown bonding mode of the diboran(4)yl ligand,
that is, an unexpected additional dative Pt B bond to the
ꢀ
valent transition-metal complex, consists of the oxidative
[1]
ꢀ
addition of the B B bond to generate cis-bis(boryl) species.
For instance, the oxidative addition of tetraalkoxydibor-
anes(4) to Pd0 or Pt0 centers has been widely applied in
organic chemistry to prepare diverse substrates for Suzuki–
ꢀ
Miyaura type coupling reactions.[1] In addition to B B bond
ꢀ
ꢀ
cleavage, halide substituted diboranes(4) potentially offer the
opportunity for oxidative addition of a boron–halide bond
eventually resulting in the formation of transition-metal
diboran(4)yl species. The latter process was shown to proceed
readily for a variety of substituted boron halides XBR2, and
has been used extensively in the preparation of transition-
metal boryl complexes.[2] By contrast, related studies on the
reactivity of halide substituted diboranes(4) still remain
scarce, which is somewhat surprising regarding the afore-
mentioned high significance of diboranes(4) as reagents in
organic chemistry. Only a single relevant study has appeared
in the literature so far. In 1999 the group of Norman studied
the reactivity of B2(NMe2)2Cl2 towards [(h2-C2H4)Pt(PPh3)2].
In this case however, no evidence for the oxidative addition of
second boron center, thus combining two types of electron-
precise Pt B interaction in one molecule. In the early days,
ꢀ
dative metal boron bonds played an important role in the
development of the concept of transition-metal basicity due
to the electron deficiency of the boron center.[1f,2d,7] Even
though these kind of interactions had already been postulated
in the 1960s, it was not until 1999 that the first example of a
ꢀ
supported dative M B bond was structurally substantiated
for the ruthenium boratrane [HRu{B(mt)3}(PPh3)] (mt =
2-sulfanyl-1-methylimidazole),[8] while structural evidence
ꢀ
for unsupported dative M B bonds still remains elusive.
During the last decade a large number of supported species
have been structurally characterized containing a variety of
different metal centers and donor arm architectures.[2d,7c–e] For
platinum however, the number of known examples is rather
limited. To date, only a couple of platinaboratrane species
have been published, in which the boron center is part of a
rigid tris(methimazolyl)borate (k4-SSSB, A),[9] diphosphanyl-
borane (k3-PPB, B)[10] or trisphosphanylborane (k4-PPPB,
ꢀ
the B Cl bond could be deduced from the experimental
ꢀ
results, and exclusively products arising from B B bond
activation were verified, that is, cis-[(Ph3P)2Pt{B-
(NMe2)(Cl)}2] and trans-[(Ph3P)2Pt(Cl){B(NMe2)(Cl)}].[3] An
alternative approach towards transition-metal diboran(4)yl
complexes has been developed in our group. We demon-
strated that the diboran(4)yl moiety can be readily introduced
by salt elimination reactions of B2(NMe2)X2 (X = Cl,[4] Br,[5]
I[6]) with anionic half-sandwich carbonyl species M’-
[CpM(CO)n] (M’ = Na, K; M = Mo, W, Fe, Ru).
Owing to our longstanding interest in transition-metal
boron chemistry, we recently began to reconsider the
reactivity of halide-substituted diboranes(4) towards low-
valent platinum complexes and to evaluate the question,
C)[11] cage structure. The strength of the dative Pt B
ꢀ
[*] Prof. Dr. H. Braunschweig, Dipl.-Chem. A. Damme, Dr. T. Kupfer
Institut fꢀr Anorganische Chemie, Julius-Maximilians-Universitꢁt
Wꢀrzburg
ꢀ
interaction, as indicated by the Pt B bond lengths (2.1–
2.5 ꢀ), is found within a large range depending on the ligand
framework. However, the rigid ligand design forces the boron
center into geometric proximity to the metal center, thus
facilitating the dative interaction. This feature is in stark
contrast to the dative Pt B interaction in 1 (see below), in
which the flexibility of the pendant diboran(4)yl substituent is
at least in principle retained.
Am Hubland, 97074 Wꢀrzburg (Germany)
E-mail: h.braunschweig@mail.uni-wuerzburg.de
Homepage: http://www-anorganik.chemie.uni-wuerzburg.de/
Braunschweig/
ꢀ
[**] This work was supported by the Deutsche Forschungsgemeinschaft.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201102593.
Angew. Chem. Int. Ed. 2011, 50, 7179 –7182
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7179

Communications
The reaction of [Pt(PEt3)3] (3) with two equivalents of
B2Mes2Br2 (4) results in the selective oxidative addition of
one B Br bond to the Pt0 center to afford trans-
ꢀ
[(Et3P)2Pt(Br){B(Mes)B(Mes)(Br)}] (1), as judged by multi-
nuclear NMR spectroscopy of the reaction mixture
(Scheme 1).[12] However, the reaction is accompanied by the
Scheme 1. Synthesis of 1 by selective oxidative addition.
Figure 1. Molecular structure of 1 (left) and 2 (right) in the solid state.
Hydrogen atoms are omitted for clarity.
release of one equivalent of PEt3 to generate the reactive
platinum species {Pt(PEt3)2}. Free PEt3 subsequently reacts
with 4 to form a phosphine–diborane(4) adduct, for which
reason this transformation requires the presence of two
equivalents of 4 to go to completion. The oxidative addition of
[11b]
ꢀ
which a dative Pt B interaction has been verified.
The
platinum center displays a distorted square-pyramidal geom-
etry with the boron atom B2 involved in the dative bonding in
the apical position. While the basal atoms Pt1, Br1, P1, B1,
and P2 still form an almost regular square base (ꢀPt = 360.18),
the apical B2 deviates significantly from its ideal position (B1-
Pt1-B2 40.52(10)8; B2-Pt1-Br1 138.64(6)8), which has to be
ascribed to the rigidity of the diboran(4)yl ligand. The
perpendicular orientation of the PtBB triangle with respect
to the square base (P1-Pt1-B1-B2 97.02(13)8; P2-Pt1-B1-B2
ꢀ91.12(13)8) is also noteworthy. The structural parameters of
the diboran(4)yl unit are only slightly altered upon oxidative
ꢀ
the B Br bond occurs spontaneously and conversion to 1 is
quantitative within seconds. After workup, 1 was isolated as a
red-orange material in poor yields of 17%. Higher yields were
thwarted by the rather difficult separation of the phosphine–
diborane(4) side product by fractional crystallization.
The identity of 1 was unambiguously ascertained by NMR
spectroscopy, elemental analysis, and X-ray diffraction.[12] The
1H NMR spectrum of 1 in solution (C6D6) features well-
separated sets of signals for two chemically nonequivalent
PEt3 groups (d = 0.75, 1.52 ppm and d = 0.92, 2.10 ppm).
Similarly, the 13C NMR spectrum shows two distinct reso-
nances at d = 9.29 and 9.12 ppm, respectively, for the methyl
groups of the PEt3 ligands, while the assignments were further
verified by 2D NMR experiments. In agreement with these
findings, the 31P NMR spectrum of 1 displays a signal pattern
indicative of an ABX spin system. Thus, two doublets are
observed at d = 5.32 and 8.56 ppm with coupling constants of
1JP–Pt = 2919 Hz (2JP–P = 333 Hz) and 2862 Hz (2JP–P = 333 Hz),
respectively. As expected, two signals are observed in the
11B NMR spectrum of 1 at d = 56 and 108 ppm, which appear
either high-field or low-field shifted with respect to those of
B2Mes2Br2 (d = 86 ppm).[13] The chemical shifts of both
resonances are fully consistent with the molecular structure
of 1 in the solid state (see below). While the signal at d =
108 ppm substantiates the presence of a metal-bound boryl
moiety, the pronounced high-field shift of the second
addition to the Pt0 center (B1 B2 1.649(4) ꢀ, B2 Br2
ꢀ
ꢀ
[13]
ꢀ
ꢀ
2.027(3) ꢀ; cf. 4: B B 1.65 ꢀ, B Br 1.93 ꢀ). The trans-
influence of the diboran(4)yl ligand in 1 can be estimated by
comparison of relevant crystal structure data to those of the
related monoboryl species trans-[(Cy3P)2Pt(Br){B(Br)(R)}]
[R = NMe2 (5), Mes (6), tBu (7)].[2b] Thus, the Pt1 Br1
ꢀ
ꢀ
(2.6210(4) ꢀ) and Pt1 B1 (2.038(3) ꢀ) bond lengths in 1,
ꢀ
ꢀ
which are similar to those in 5–7 (Pt Br 2.61–2.65 ꢀ; Pt B
1.98–2.01 ꢀ), suggest that the diboran(4)yl moiety also exerts
a strong trans-influence.
ꢀ
Further support for the presence of a dative Pt B bonding
interaction is deduced from DFT calculations.[12] For this
purpose, in addition to single-point calculations on 1 using
crystal structure coordinates (1CS), a geometry optimization
without symmetry restraints (1OPT) was conducted at the
B3LYP level of theory. To further evaluate the strength of the
ꢀ
dative Pt B bond, we also studied two additional model
ꢀ
resonance (d = 56 ppm) already indicates a dative Pt B
interaction even in solution.
species, that is, 8 and 9 (Figure 2), in which any dative
interactions are suppressed. In 8, the mesityl ligand at B2 is
substituted by a NMe2 functionality, which results in the
The most intriguing feature of the molecular structure of 1
ꢀ
=
in the solid state (Figure 1) is the verification of the dative Pt
occupation of the empty pz orbital at B2 by a B N p bond
B bonding to the second boron center (Pt1!B2), which is
without significantly altering the boron environment. In
addition, we chose the Lewis acid–base adduct of 1 with
HCN (9) to account for both structural and electronic effects
ꢀ
nicely illustrated by the short Pt1 B2 distance (2.531(3) ꢀ)
and by the small values found for the bond angles Pt1-B1-B2
ꢀ
(86.05(15)8) and B1-Pt1-B2 (40.52(10)8). In comparison to the
on the Pt1 B2 interaction upon changing the hybridization of
Pt1 B1 bond (2.038(3) ꢀ), which is found in a typical range
B2 from sp2 to sp3, which renders any dative interactions
ꢀ
ꢀ
ꢀ
for platinum boryl species, the Pt1 B2 bond (2.531(3) ꢀ) is
impossible. The relative strength of the dative Pt1 B2 bond in
1 was subsequently assessed by comparison of the relevant
Wiberg bond indices (WBI, Table 1). While the WBIs of the
Pt1 B1 bond are of similar magnitude for all complexes
elongated by approximately 25%. However, this bond length
is comparable to that observed for the k3-PPB platinabora-
trane [PtCl2(dpb)] (dpb = PhB[2-(PiPr2)C6H4]2, type B), in
ꢀ
7180
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 7179 –7182

1H, 13C, and 31P NMR spectroscopic parameters of 2 in
solution are unremarkable and fully consistent with the
anticipated structural motif. Conclusive evidence for the
ꢀ
absence of a dative Pt1 B2 interaction is provided by
11B NMR spectroscopy. While the two 11B NMR resonances
of 1 have been detected either high-field (B2) or low-field
(B1) shifted with respect to the diborane(4) precursor, both
11B NMR signals of 2 (B2: d = 41 ppm; B1 d = 51 ppm) are
shifted to lower field by 3 and 13 ppm in comparison to those
of 11 (d = 38 ppm).[14] Accordingly, the characteristic high-
field shift of the 11B NMR resonance of B2, indicative of the
Figure 2. Natural charges of complexes 1CS, 1OPT, 8, and 9.
ꢀ
presence of a dative Pt1 B2 bond in 1, is not observed for 2,
which suggests that such an interaction is indeed absent.
This was confirmed by X-ray diffraction analysis on single
crystals of 2.[12] As depicted in Figure 1, the molecular
structure of 2 in the solid state features a square-planar
platinum center (ꢀPt = 360.78) without any evidence for a
Table 1: Calculated relevant WBIs for complexes 1CS, 1OPT, 8, and 9.
ꢀ
ꢀ
ꢀ
Pt1 B1
Pt1 B2
B1 B2
1CS
1OPT
8
0.6163
0.6730
0.7566
0.7830
0.2384
0.1652
0.0404
0.0214
1.1457
1.1082
1.0156
0.9168
ꢀ
dative bonding interaction between Pt1 and B2. The Pt1 B2
distance of 3.298 ꢀ is considerably elongated with respect to
that found in 1 (2.531(3) ꢀ), and dative interactions appear
very unlikely. This is also evidenced by the angle B2-B1-Pt1
(119.4(3)8), which is noticeably widened compared to that
observed in 1. However, the Pt-B-B triangle is again oriented
almost perpendicular to the plane defined by Pt1, P1, P2, B1,
and Br1 (P1-Pt1-B1-B2 92.7(3)8; P2-Pt1-B1-B2 ꢀ80.5(3)8).
Other parameters involving Pt1 are unremarkable, and bond
lengths and angles are found in the typical range for trans-
bisphosphineplatinum(II) species.[2b] Within the diboran(4)yl
9
ꢀ
studies, the WBIs of the Pt1 B2 bond significantly differ for
the species 1CS/1OPT (0.2384/0.1652) and 8/9 (0.0404/0.0214),
respectively. These findings clearly indicate the presence of
ꢀ
noticeable dative Pt1 B2 interactions in 1CS and 1OPT. As
expected, this interaction is much more pronounced in the
solid state (1CS) than in the gas phase (1OPT).
ꢀ
moiety, the elongated B B distance (1.742(8) ꢀ; cf. 1
Determination of the natural charges (Figure 2) fully
confirms these assumptions. Consistent with the stronger
1.649(4) ꢀ) is noteworthy, which is clearly associated with
the presence of NMe2 p donor functionalities at boron.
In conclusion, we have disclosed a novel bonding mode of
the diboran(4)yl ligand consisting of two different electron-
ꢀ
dative Pt1 B2 bond in 1CS as compared to 1OPT, the boron
atom B2 bears considerably less positive charge in 1CS
(+0.19092; cf. 1OPT + 0.26975), indicating enhanced electron
donation from the platinum center to the electron-deficient
boron atom B2. With regard to the model compounds 8 and 9,
ꢀ
precise Pt B interactions (boryl/dative). For this, the com-
monly observed reactivity of diborane(4) derivatives towards
ꢀ
low-valent platinum species, that is, the activation of the B B
ꢀ
the absence of any dative Pt1 B2 interaction is visualized by
bond, has been successfully suppressed in favor of a selective
oxidative addition of the boron–halide bond in 4 and 11. This
approach is without precedent in transition-metal boron
chemistry and has enabled the isolation and full character-
ization of the diboran(4)yl complexes 1 and 2. Structural
analysis of 1 revealed the presence of an uncommon and
the strongly diminished positive charge on Pt (8 + 0.10591, 9
+ 0.09847; cf. 1CS + 0.13934). Accordingly, the platinum atoms
in 8 and 9 are not able to release electron density to the boron
center B2, whose formerly empty pz orbital (1) is already
filled with electrons.
ꢀ
To further substantiate the results of the theoretical
calculations, we sought to synthesize of a diboran(4)yl species
similar to 8, for which theory predicts the absence of a dative
unexpected dative Pt B bonding interaction to the second
boron center of the diboran(4)yl ligand, which has already
been indicated by NMR spectroscopy in solution. The relative
strength of this dative bond has been evaluated by DFT
calculations, which also predicted that the introduction of p
donor functionalities at boron would eliminate such a dative
ꢀ
Pt1 B2 bond. The reaction of [Pt(PiPr3)2] (10) with one
equivalent of B2(NMe2)2Br2 (11) afforded the envisaged
complex trans-[(iPr3P)2Pt(Br){B(NMe2)B(NMe2)(Br)}] (2) in
72% yield (Scheme 2).[12] The conversion readily proceeded
ꢀ
Pt B interaction. These findings have been confirmed
experimentally with the synthesis of 2, in which no evidence
ꢀ
at room temperature and the oxidative addition of one B Br
bond to the Pt0 center was found to be highly selective
according to NMR spectroscopy of the reaction mixture.
for a dative Pt B bond was obtained. The extension of this
ꢀ
novel reactivity to other Pt0 species and other halide-
substituted diborane(4) derivatives is currently being inves-
tigated extensively in our laboratories.
Received: April 14, 2011
Published online: June 16, 2011
Scheme 2. Reaction of 10 with B2(NMe2)2Br2.
Keywords: boron · diboran(4)yl ligands · diborane(4) · platinum
.
Angew. Chem. Int. Ed. 2011, 50, 7179 –7182
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7181

Communications
[7] See, for example: a) H. Werner, Pure Appl. Chem. 1982, 54,
177 – 188; b) H. Braunschweig, Angew. Chem. 1998, 110, 1882 –
1898; Angew. Chem. Int. Ed. 1998, 37, 1786 – 1801; c) H.
Braunschweig, C. Kollann, D. Rais, Angew. Chem. 2006, 118,
5380 – 5400; Angew. Chem. Int. Ed. 2006, 45, 5254 – 5274; d) A.
Amgoune, D. Bourissou, Chem. Commun. 2011, 47, 859 – 871;
e) H. Braunschweig, R. D. Dewhurst, Dalton Trans. 2011, 40,
549 – 558.
[8] A. F. Hill, G. R. Owen, A. J. P. White, D. J. Williams, Angew.
Chem. 1999, 111, 2920 – 2923; Angew. Chem. Int. Ed. 1999, 38,
2759 – 2761.
[9] a) I. R. Crossley, A. F. Hill, Organometallics 2004, 23, 5656 –
5658; b) I. R. Crossley, A. F. Hill, Dalton Trans. 2008, 201 –
203; c) I. R. Crossley, A. F. Hill, A. C. Willis, Organometallics
2008, 27, 312 – 315.
[10] S. Bontemps, M. Sircoglou, G. Bouhadir, H. Puschmann, J. A. K.
Howard, P. W. Dyer, K. Miqueu, D. Bourissou, Chem. Eur. J.
2008, 14, 731 – 740.
[11] a) S. Bontemps, G. Bouhadir, W. Gu, M. Mercy, C.-H. Chen,
B. M. Foxman, L. Maron, O. V. Ozerov, D. Bourissou, Angew.
Chem. 2008, 120, 1503 – 1506; Angew. Chem. Int. Ed. 2008, 47,
1481 – 1484; b) M. Sircoglou, S. Bontemps, G. Bouhadir, N.
Saffron, K. Miqueu, W. Gu, M. Mercy, C.-H. Chen, B. M.
Foxman, L. Maron, O. V. Ozerov, D. Bourissou, J. Am. Chem.
Soc. 2008, 130, 16729 – 16738.
[12] Experimental procedures, details on the X-ray diffraction and
theoretical studies, as well as analytical data of all compounds
are provided in the Supporting Information. CCDC-821339 (1)
and CCDC-821340 (2) contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[1] See, for example: a) T. Ishiyama, N. Matsuda, N. Miyaura, A.
Suzuki, J. Am. Chem. Soc. 1993, 115, 11018 – 11019; b) T.
Ishiyama, N. Matsuda, M. Murata, F. Ozawa, N. Miyaura,
Organometallics 1996, 15, 713 – 720; c) T. B. Marder, N. C.
Norman, Top. Catal. 1998, 5, 63 – 73; d) G. J. Irvine, M. J. G.
Lesley, T. B. Marder, N. C. Norman, C. R. Rice, E. G. Robins,
W. R. Roper, G. R. Whittell, L. J. Wright, Chem. Rev. 1998, 98,
2685 – 2722; e) T. Ishiyama, N. Miyaura, J. Organomet. Chem.
2000, 611, 392 – 402; f) H. Braunschweig, M. Colling, Coord.
Chem. Rev. 2001, 223, 1 – 51; g) H. Braunschweig, M. Lutz, K.
Radacki, Angew. Chem. 2005, 117, 5792 – 5796; Angew. Chem.
Int. Ed. 2005, 44, 5647 – 5651; h) H. Braunschweig, T. Kupfer, M.
Lutz, K. Radacki, F. Seeler, R. Sigritz, Angew. Chem. 2006, 118,
8217 – 8220; Angew. Chem. Int. Ed. 2006, 45, 8048 – 8051; i) H.
Braunschweig, T. Kupfer, J. Am. Chem. Soc. 2008, 130, 4242 –
4243; j) H. Braunschweig, M. Kaupp, C. J. Adams, T. Kupfer, K.
Radacki, S. Schinzel, J. Am. Chem. Soc. 2008, 130, 11376 –
11393.
[2] See, for example: a) H. Braunschweig, K. Radacki, D. Rais, K.
Uttinger, Angew. Chem. 2005, 117, 5796 – 5799; Angew. Chem.
Int. Ed. 2005, 44, 5651 – 5654; b) H. Braunschweig, P. Brenner,
A. Mꢁller, K. Radacki, D. Rais, K. Uttinger, Chem. Eur. J. 2007,
13, 7171 – 7176; c) H. Braunschweig, K. Radacki, A. Schneider,
Science 2010, 328, 345 – 347; d) H. Braunschweig, R. D. Dew-
hurst, A. Schneider, Chem. Rev. 2010, 110, 3924 – 3957.
[3] D. Curtis, M. J. G. Lesley, N. C. Norman, A. G. Orpen, J.
Starbuck, J. Chem. Soc. Dalton Trans. 1999, 1687 – 1694.
[4] H. Braunschweig, B. Ganter, M. Koster, T. Wagner, Chem. Ber.
1996, 129, 1099 – 1101.
[5] H. Braunschweig, M. Koster, R. Wang, Inorg. Chem. 1999, 38,
415 – 416.
[6] a) H. Braunschweig, M. Koster, K. W. Klinkhammer, Angew.
Chem. 1999, 111, 2368 – 2370; Angew. Chem. Int. Ed. 1999, 38,
2229 – 2231; b) H. Braunschweig, K. W. Klinkhammer, M.
Koster, K. Radacki, Chem. Eur. J. 2003, 9, 1303 – 1309.
[13] H. Hommer, H. Nꢂth, J. Knizek, W. Ponikwar, H. Schwenk-
Kircher, Eur. J. Inorg. Chem. 1998, 1519 – 1527.
[14] A. Moezzi, M. M. Olmstead, P. P. Power, J. Chem. Soc. Dalton
Trans. 1992, 2429 – 2434.
7182
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 7179 –7182